In recent years, devices that digitally treat image information, and perform compression coding of an image, by employing an encoding system that compresses the image by orthogonal transform such as discrete cosine transform, and motion compensation, using redundancy specific to the image information, with a view to highly efficient transmission and accumulation of information, have been becoming popular. Examples of the encoding system include moving picture experts group (MPEG), H.264, and MPEG-4 Part 10 (advanced video coding, hereinafter, described as AVC).
Currently, aiming at further improvement of encoding efficiency from H.264/AVC, standardization of the encoding system called high efficiency video coding (HEVC) has been pushed forward by the joint collaboration team—video coding (JCTVC) that is a joint standardization organization of ITU-T and ISO/IEC. Further, as its extension, standardization of encoding about multi-view image has been carried out in parallel, and Non-Patent Document 1 has been issued as a draft about the encoding of multi-view image.
By the way, a reference list, which is a list in which candidates of reference images for decoding a B picture are listed in a priority order, is generated from a head of reference images listed in a temporal list in a priority order.
In Non-Patent Document 1, whether a long-term reference image and a short-term reference image are added to the temporal list can be switched by a slice parameter set (SPS) or a slice header.
In contrast, reference relationship of inter-layers (views) is defined in a video parameter set (VPS). A layer, reference of which is specified by the VPS, is always added to the temporal list. That is, the reference image of the inter-layer can be controlled only with the VPS transmitted only once with a sequence.
Therefore, in Non-Patent Document 2, supplemental enhancement information (SEI) for changing the reference relationship of views has been proposed.